Cone-beam computed tomography (CBCT) has been broadly used in image guided radiation therapy (IGRT) and adaptive radiation therapy (ART), to acquire the updated patient's geometry for precise targeting and treatment adaptation. However, the repeated use of CBCT during a treatment course has raised a serious concern on excessive x-ray imaging doses delivered to patients, which has greatly limited the maximal exploitation of the potential of modern radiotherapy. Especially for pediatric patients, this concern has prohibited the use of IGRT and ART, leading to compromised treatment outcome. Advanced iterative reconstruction algorithms, based on compressed sensing techniques, have demonstrated tremendous power in reconstructing CBCT images from very few and/or noisy projections, resulting in dramatically reduced imaging dose. However, these algorithms are very computationally inefficient and thus cannot be used in most clinical applications. We have recently made a breakthrough in developing an innovative CBCT reconstruction algorithm with a mathematical structure perfect for parallelization on a graphics processing unit (GPU) platform. Our preliminary results have shown that we can improve the efficiency by a factor of 100 over existing iterative algorithms and reduce the imaging dose by factor of 40~100 compared to the current clinical standard. Our goal is to develop this promising algorithm into a clinically functioning CBCT reconstruction system which can produce high quality CBCT images at extremely low radiation dose (<1% of the current dose) and high speed (< 5 seconds), by pursuing the following two specific aims: SA1. We will develop a GPU-based system to reconstruct high quality CBCT images at ultra-low radiation dose and ultra-high speed. SA2. We will evaluate the system through a series of numerical, phantom, and patient studies, demonstrate the gain in imaging dose reduction, and establish clinical protocols under various clinical conditions. Upon the completion of the proposed project, a clinically ready-to-use CBCT reconstruction system with ultra-low dose and ultra-fast performance will have been systematically developed and evaluated. Clinical introduction of such a system will significantly benefit a large number of patients receiving modern radiotherapy. Especially, our work will for the first time make IGRT and ART clinically available for pediatric patients. PUBLIC HEALTH RELEVANCE: This project is to develop an ultra fast and extremely low dose cone beam CT (CBCT) reconstruction system for image guided adaptive radiotherapy. Specifically, we will develop innovative CBCT reconstruction algorithms that can reduce the imaging dose to less than one percent of the current state of the art. More importantly, the mathematical structure of the new algorithms is perfect for GPU parallelization which makes the fast reconstruction clinically feasible.